Fig. 10
From: Evaluating the authenticity of ChatGPT responses: a study on text-matching capabilities
The capability of ChatGPT (model 4) to generate responses with less than25% matching (Overall)
Fig. 10 | International Journal for Educational Integrity